AU Monocerotis: Interacting Binary Dynamics

• AU Monocerotis is an eclipsing, interacting binary system with a hotter gainer and an evolved donor transferring mass via Roche-lobe overflow.
• Detailed modeling from photometry and spectroscopy has yielded precise orbital and stellar parameters, evidencing a persistent circumprimary accretion disk with double-peaked Hα emission.
• Recent studies show that the system’s long photometric cycle vanished around 2010, suggesting non-conservative mass transfer and challenging traditional dynamo-driven models.

Searching arXiv for recent and foundational AU Monocerotis papers to ground the article. AU Monocerotis (AU Mon; HD 50846, HIP 33237) is a Galactic eclipsing, double-lined spectroscopic Algol-type interacting binary and a member of the double periodic variable (DPV) class. It consists of a hotter, more massive gainer and a cooler, evolved donor that fills its Roche lobe and transfers mass to the gainer. The system has an orbital period of about $11.113$ d and exhibits persistent circumstellar structure, most prominently a circumprimary accretion disk signaled by longstanding double-peaked H$\alpha$ emission. Historically, AU Mon also showed a long photometric cycle near $411$–$417$ d, a defining DPV characteristic, although recent photometric analyses indicate that this long cycle vanished around 2010 while the orbital period remained essentially constant (Djurasevic et al., 2010, Celedón et al., 16 Jul 2025).

1. Classification and binary architecture

AU Mon is described as a long-period Algol-type eclipsing binary, a hot, massive Algol, and a semidetached interacting system. In this configuration, the evolved secondary or donor fills its Roche lobe and transfers matter to the hotter primary or gainer. The orbit is treated as circular in the light-curve analysis, and the accepted orbital period from CoRoT-based work is $P_\mathrm{o}=11.1130374(1)$ d (Djurasevic et al., 2010).

Disk-inclusive photometric modeling fixes several binary parameters from spectroscopy and prior work: $q=M_c/M_h=0.17$, donor temperature $T_c=5750\,\mathrm{K}$, gainer temperature $T_h=15890\,\mathrm{K}$, gainer non-synchronous rotation $f_h=5.2$, donor synchronous rotation $f_c=1.0$, and donor Roche-lobe filling factor $\alpha$0. The mean CoRoT solution yields $\alpha$1, $\alpha$2, $\alpha$3, $\alpha$4, $\alpha$5, and $\alpha$6. The same analysis finds that a disk model fits the light curve much better than a simple semidetached Roche model without a disk, and it also provides a more satisfactory account of the Rossiter–McLaughlin effect (Djurasevic et al., 2010).

The basic system classification is therefore not merely taxonomic. It encodes a specific dynamical state: Roche-lobe overflow from an evolved donor, accretion onto the hotter star, and persistent circumstellar matter whose radiative and dynamical effects are observable in both photometry and spectroscopy.

2. Accretion disk and circumstellar environment

The most persistent signature of AU Mon’s circumstellar environment is double-peaked H$\alpha$7 emission. In the spectroscopic accretion-structure analysis, this is treated as evidence for a persistent accretion disk around the gainer. Shellspec modeling, using optical spectra and archival IUE ultraviolet data, adopts Roche-geometry stars plus a circular, symmetric accretion disk, a straight cylindrical gas stream, and optional localized structures. The preferred spectroscopic disk has inner/outer radius $\alpha$8–$\alpha$9 and $411$0, thickness $411$1, density $411$2, maximum temperature $411$3, and turbulence $411$4. The associated gas stream has temperature $411$5, initial density $411$6, radius $411$7, initial/final velocities $411$8, and turbulence $411$9 (Atwood-Stone et al., 2012).

Photometric disk modeling based on CoRoT and $417$0-band light curves uses a geometrically and optically thick disk around the gainer, with edge approximated by a cylindrical surface and thickness varying linearly with radius. In this framework, AU Mon favors a moderately concave disk. The mean solution gives disk dimension factor $417$1, disk radius $417$2, edge and center thicknesses $417$3 and $417$4, edge temperature $417$5, and temperature exponent $417$6. The model also requires one hot spot and two bright spots on the disk edge; the hot spot has $417$7, $417$8, and $417$9, while the bright spots are interpreted as spiral-arm-like structure or other non-axisymmetric disk distortions (Djurasevic et al., 2010).

The differing disk dimensions reported by spectroscopy and continuum light-curve modeling are not treated as a formal contradiction. The spectroscopic study explicitly notes that emission-line spectroscopy is sensitive to a wavelength-dependent effective size, that earlier light-curve work modeled a non-transparent continuum disk with hot and bright spots, and that disk properties may have changed between observing epochs (Atwood-Stone et al., 2012). This suggests that “the AU Mon disk” is observationally model-dependent: its inferred size and thermal structure depend on whether the diagnostic is continuum obscuration, line emission, or ultraviolet absorption.

3. Spectroscopic diagnostics, tomography, and short-timescale variability

AU Mon has been observed with 95 high-resolution optical spectra spanning 20 years and 43 archival IUE ultraviolet spectra spanning 16 years. These data show that the accretion environment varies with orbital phase, with long-cycle phase, and on additional multi-week and multi-year timescales. In H$P_\mathrm{o}=11.1130374(1)$0, the blue and red emission peaks arise naturally from the approaching and receding sides of the rotating disk, while the central absorption is shaped by eclipse geometry and by extra absorbing material in the stream or envelope. The central absorption is reported to be deeper for $P_\mathrm{o}=11.1130374(1)$1, which is interpreted as evidence for long-period changes in the accretion environment, especially a thicker or more opaque disk or envelope during that interval (Atwood-Stone et al., 2012).

H$P_\mathrm{o}=11.1130374(1)$2 Doppler tomography provides an independent test of the spectroscopic model. Sequential subtraction of the stars, disk, stream, and a localized region shows that the strongest H$P_\mathrm{o}=11.1130374(1)$3 emission is concentrated where a Keplerian disk should appear in velocity space, confirming that the disk dominates the line emission. The gas stream contributes a weaker but visible component, mostly at velocities below $P_\mathrm{o}=11.1130374(1)$4. Even after subtracting the modeled components, residual emission remains at sub-Keplerian velocities. The spectroscopic study interprets this residual as evidence that the observed gas distribution is not fully described by a simple symmetric disk plus stream; possibilities raised in the paper include an asymmetric disk, gas continuing beyond the stream-impact region and slowing down, or a vertical outflow or jet-like component (Atwood-Stone et al., 2012).

A transient ultraviolet absorption feature, seen mainly in Si IV and occasionally in Al III and Si II, adds further complexity. It appears only near $P_\mathrm{o}=11.1130374(1)$5 to $P_\mathrm{o}=11.1130374(1)$6, is not obviously tied to long-period phase, and often requires blueshifted absorption. Shellspec tests show that it can be modeled by adding an extra outflow spot near the disk-stream interaction region, leading to the interpretation of an occasional outflow or wind launched from the vicinity of the impact site (Atwood-Stone et al., 2012).

Short-timescale photometric residuals also exist after subtraction of the binary-plus-disk model. Fourier analysis of the CoRoT residuals yields two stable frequencies, $P_\mathrm{o}=11.1130374(1)$7 and $P_\mathrm{o}=11.1130374(1)$8, consistent with frequencies previously reported from the same photometric material. The fast variation near primary eclipse is suggested to arise from a hidden oscillation mode of the gainer rather than from the donor or the disk (Djurasevic et al., 2010).

4. Evolutionary state and mass-exchange history

The evolutionary interpretation of AU Mon has been addressed through a multi-parameter $P_\mathrm{o}=11.1130374(1)$9 minimization between observed system parameters and a grid of 561 conservative and non-conservative binary evolutionary tracks from Van Rensbergen et al. The best-fit solution places the system at an age of about $q=M_c/M_h=0.17$0 yr and identifies the present state as a Case-B mass-exchange episode. The inferred initial configuration is $q=M_c/M_h=0.17$1 and $q=M_c/M_h=0.17$2 in a 3.0-day orbit, evolving to a current donor of about $q=M_c/M_h=0.17$3 and gainer of about $q=M_c/M_h=0.17$4 (Mennickent, 2014).

In this evolutionary picture, the donor is an evolved star whose hydrogen core is completely exhausted. The donor has therefore already passed through core hydrogen burning and is now transferring mass after leaving the main sequence, which is the defining condition for Case-B Roche-lobe overflow. The paper emphasizes that this stage is consistent with the reported semidetached geometry and with the existence of a circumprimary accretion disk. It also notes that AU Mon is observed only about $q=M_c/M_h=0.17$5 years after the beginning of the mass-transfer episode, making the current transfer stage extremely short compared with the full $q=M_c/M_h=0.17$6 Myr system age (Mennickent, 2014).

This evolutionary reconstruction underpins a broader interpretation of AU Mon as an evolved semidetached binary whose current appearance is the product of long-lived binary evolution and a comparatively recent, active Roche-lobe overflow phase. The evolutionary model is conservative in its best fit, but the paper’s own discussion shows that strictly conservative transfer cannot by itself explain all current observables.

5. Mass-transfer rate, orbital-period stability, and non-conservative loss

A central problem in AU Mon is the tension between evolutionary and orbital-period diagnostics. The best-fit conservative evolutionary model implies a donor mass-transfer rate

$q=M_c/M_h=0.17$7

with the gainer accreting at essentially the same rate in the long term. For conservative transfer, Huang’s period relation,

$q=M_c/M_h=0.17$8

gives $q=M_c/M_h=0.17$9, corresponding to about $T_c=5750\,\mathrm{K}$0 s yr$T_c=5750\,\mathrm{K}$1. That is far larger than the observational upper limit from the photometric baseline considered in the evolutionary study (Mennickent, 2014).

Long-baseline photometric timing strengthens this discrepancy. An O–C analysis over 46.3 years yields an orbital-period change of only $T_c=5750\,\mathrm{K}$2 s yr$T_c=5750\,\mathrm{K}$3, consistent with essentially constant orbital period over recent decades. Under a fully conservative mass-transfer regime, this constrains the characteristic transfer rate to $T_c=5750\,\mathrm{K}$4. The same analysis notes that if the system is non-conservative, a larger $T_c=5750\,\mathrm{K}$5 can be accommodated, with an upper compatible extreme of $T_c=5750\,\mathrm{K}$6 (Celedón et al., 16 Jul 2025).

The preferred resolution is therefore a recent departure from conservativeness. The evolutionary study estimates a timescale

$T_c=5750\,\mathrm{K}$7

using $T_c=5750\,\mathrm{K}$8, and argues that only a small fraction of the system’s history needs to be non-conservative for the present stellar parameters to remain close to the conservative best-fit track while the current orbital period remains nearly unchanged. In the $T_c=5750\,\mathrm{K}$9 formalism, where

$T_h=15890\,\mathrm{K}$0

is the fraction of donor mass accreted by the gainer and $T_h=15890\,\mathrm{K}$1 parameterizes the specific angular momentum carried away by lost matter, particular pairs of $T_h=15890\,\mathrm{K}$2 can produce nearly constant $T_h=15890\,\mathrm{K}$3. If mass is lost with the specific angular momentum of the gainer, then

$T_h=15890\,\mathrm{K}$4

which is too small to keep the period constant. If instead the lost material forms or couples to a circumbinary disc, a characteristic $T_h=15890\,\mathrm{K}$5 becomes possible, and a mass-loss efficiency around $T_h=15890\,\mathrm{K}$6 can preserve an almost constant orbital period (Mennickent, 2014).

Accordingly, AU Mon is frequently interpreted not as a contradiction between evolutionary theory and timing data, but as a system whose long-term history may be mostly conservative while its present state includes substantial mass and angular-momentum loss through circumbinary material or disk-related outflows.

6. Long photometric cycle and its disappearance

As a DPV, AU Mon historically displayed a long photometric cycle in addition to its orbital variability. DPVs are semi-detached binaries with long cycles typically near $T_h=15890\,\mathrm{K}$7. AU Mon’s long cycle was first identified as a $T_h=15890\,\mathrm{K}$8–$T_h=15890\,\mathrm{K}$9 d modulation with amplitude of $f_h=5.2$0 mag in $f_h=5.2$1, and it was long interpreted as a key tracer of the system’s mass-transfer and accretion behavior. Earlier photometric modeling argued that the $f_h=5.2$2-day brightness modulation could not be explained simply by changing the disk size or the disk active regions, and therefore favored variable attenuation by circumbinary material rather than disk-geometry changes alone (Djurasevic et al., 2010).

A 46.3-year reanalysis of archival and survey photometry shows that the long cycle is not stationary. In the $f_h=5.2$3 band, the classical $f_h=5.2$4 d signal is present in early data, persists through ASAS-3, weakens continuously, and becomes visually undetectable around MJD $f_h=5.2$5, that is, around 2010. The disappearance is especially clear in $f_h=5.2$6, whereas in $f_h=5.2$7 a remnant of long-cycle-like variability remains more detectable for some interval (Celedón et al., 16 Jul 2025).

When the canonical long cycle is absent, other transient periodicities appear in redder bands. In the $f_h=5.2$8-band KWS data, the dominant orbital-subtracted signal is a period of $f_h=5.2$9 d with amplitude $f_c=1.0$0 mag, described by

$f_c=1.0$1

This modulation persists for at least $f_c=1.0$2 d and then disappears around 2020. In APASS data around 2013, another transient periodicity near $f_c=1.0$3 d is detected, strongest in the redder $f_c=1.0$4 filter, with $f_c=1.0$5 mag. The chronology reported in the recent study is therefore: classical $f_c=1.0$6 d cycle in early data, weakening and disappearance around 2010 in $f_c=1.0$7, appearance of transient longer and shorter red-band modulations, and disappearance of those transients by about 2020 (Celedón et al., 16 Jul 2025).

The disappearance of the AU Mon long cycle is presented as the second documented sudden vanishing of a DPV long cycle after TYC 5353-1137-1. The phenomenon is explicitly distinguished from a simple secular shortening of the long period, such as reported for OGLE-LMC-DPV-065. For AU Mon, the evidence supports a genuine vanishing rather than a smooth migration of the long-cycle timescale (Celedón et al., 16 Jul 2025).

For theory, this behavior is a strong constraint. In magnetic-dynamo interpretations of DPV cycles, modulation of the donor’s magnetic activity changes its quadrupole moment and the mass-transfer rate, thereby producing the long photometric cycle. The observed suppression of AU Mon’s long cycle implies that whatever parameter or parameters control the dynamo-driven modulation changed enough to suppress the cycle entirely. More generally, the coexistence of a stable orbital period with an unstable and vanishing long cycle shows that DPV phenomenology is more diverse than a fixed ratio $f_c=1.0$8 would suggest (Celedón et al., 16 Jul 2025).